Table 4 Diagnostic efficacy of different methods in training set and test set.
Set | Methods | AUC (95% CI) | P value | Cut-off | Sensitivity | Specificity | LR+ | LR- |
|---|---|---|---|---|---|---|---|---|
Training set | CEUS LI_RADS | 0.754 (0.685–0.815) | < 0.0001 | 0.272 | 64.37% | 86.46% | 4.75 | 0.41 |
STE | 0.909 (0.857–0.946) | < 0.0001 | 0.414 | 81.61% | 88.54% | 7.12 | 0.21 | |
Nomogram | 0.988(0.963–0.999) | < 0.0001 | 0.548 | 95.40% | 95.83% | 22.9 | 0.048 | |
Test set | CEUS LI_RADS | 0.824 (0.722–0.901) | < 0.0001 | 0.236 | 64.86% | 100% | — | 0.35 |
STE | 0.923 (0.840–0.971) | < 0.0001 | 0.260 | 86.49% | 90.48 | 9.08 | 0.15 | |
Nomogram | 0.978 (0.917–0.998) | < 0.0001 | 0.299 | 94.59 | 95.24 | 19.86 | 0.057 |
